Operating room scheduling An evaluation of alternative scheduling approaches to improve OR efficiency and minimize peak demands for ward beds at SKB Winterswijk

Knoeff, Thijs (2010) Operating room scheduling An evaluation of alternative scheduling approaches to improve OR efficiency and minimize peak demands for ward beds at SKB Winterswijk.

[img]
Preview
PDF
1MB
Abstract:Introduction The OR-department of Streekziekenhuis Koningin Beatrix in Winterswijk faces the need to improve efficiency, while the OR schedule causes high peak requirements for beds on surgical wards, waiting lists for surgery remain long and OR-planners deal with increasing workloads due to a multitude of equipment related constraints. The OR-department of SKB can be characterized as ‘high volume, low complexity’, with short case durations and a small number of operation types covering the majority of operations performed. Objectives The aim of this research is to develop a surgery scheduling system for the OR-department of SKB that increases OR efficiency, levels bed occupancy at the surgical wards and reduces workload for planning personnel, while satisfying the constraints set by limited resource availability (instrument sets required, ward bed capacity, and equipment required). OR efficiency is measured by two measures: (1) idle time of the OR at the end of the day, after having performed all planned surgeries, (2) overtime required for performing all planned surgeries. Inefficiencies due to idle time and overtime are equivalent to substantial costs for the hospital management and should therefore both be minimized. The outcome of this research is furthermore required to consist of directions, rules and/or procedures for surgery scheduling, rather than custom-built planning software and is required to be able to be implemented within the restrictions of current information systems as much as possible. Methods We evaluate several different scheduling approaches by using self-programmed scheduling software and evaluate the performance of our schedules by testing these in a couple of event-based simulation runs. We run these tests on modelled data, which we derive from actual historic data from the hospital information systems. Model validation shows that we may assume the results of our study to be sufficiently valid for the real life situation at SKB, within the context and assumptions of our research. The scheduling approaches we test consist of a combination of scheduling heuristics of two sorts: constructive and improvement heuristics. The constructive heuristics resemble a structured process of efficiently filling the OR capacity with operations, while taking all constraints into account with regard to required instrument sets, required equipment, maximum waiting time of the patient and limited capacity at the surgical ward. The improvement heuristics resemble a trial-and-error process of trying to improve the schedule with regard to the performance indicators (idle time, overtime and bed Operating room scheduling in SKB Winterswijk Thijs Knoeff 6 occupancy levelling) while maintaining a feasible schedule with regard to the resources required. Furthermore, we test several planning targets to address the question at which target level planners should be aiming to optimally balance idle time and overtime. A major part of our method focuses on the use of a Master Surgical Schedule (MSS). The underlying idea of such an approach is that surgeries of some same surgery type are very similar. The effort of scheduling these surgeries could be reduced enormously by creating a cyclic blueprint, containing ‘slots’ of these surgery types. Real surgeries are then assigned to empty ‘slots’ of the corresponding surgery type. This means that, when the hospital manages to construct a feasible, acceptable and optimized master schedule (MSS), weekly planning would boil down to filling in a ‘blanks exercise’. All the constraints and performance objectives (e.g. levelled bed occupancy) are already incorporated in the MSS. The MSS approach has the promise of greatly reducing complexity at the operational offline planning level, while performance, which is based on the quality of the master schedule, may greatly improve if you manage to construct an excellent and well balanced MSS. We evaluate such an approach and vary several parameter values in order to determine the ‘optimal’ cycle length and number of slots for each surgery type. Results The simulation data show that the results are best for an approach with a combination of a straightforward (Random Fit) constructive heuristic and the most advanced form of improvement heuristic we tested (RE123+), while using a straightforward 100% planning target. Regrettably, running the improvement heuristics is not doable for a human planner, so this approach dit not meet all criteria. The best feasible approach consists of the use of a Master Surgical Schedule with cycle length of 4 weeks, and the use of straightforward Random Fit and 100% planning target for the remaining surgeries. This approach leads to a reduction of overtime and idle time of respectively 46% and 34%, while reducing fluctuation in bed occupancy levelling by a mere 56% on average. Furthermore, over 83% of all surgeries can be scheduled within the ‘slots’ of the MSS, greatly reducing the complex puzzle that planners need to solve each week. Recommendations SKB is recommended to: - define and maintain surgery types and use these for in OR planning - use predictions based on historical data for operation duration and turnover time for each surgery type, rather than surgeon-based estimates - construct a MSS consisting of an agreed number of slots for each surgery type - use an optimized MSS to further fine-tune wishes of the relevant stakeholders in the hospital with regard to OR planning
Item Type:Essay (Master)
Clients:
SKB
Faculty:BMS: Behavioural, Management and Social Sciences
Subject:85 business administration, organizational science
Programme:Industrial Engineering and Management MSc (60029)
Link to this item:http://purl.utwente.nl/essays/60822
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page